Leon Furze

A lot of the rhetoric around artificial intelligence in education centres on the idea that AI will inevitably lead to skills loss and that this is an inherently bad thing. Like all hype at both ends of the AI spectrum, both positive and negative, this is an oversimplification, and I think we need to push back against this narrative. The AI = skills loss argument does us a disservice in two ways.

Firstly, it diminishes the role of the teacher to a “deliverer” of content, a pseudo-expert with a bag of tricks that AI now threatens. For example, saying that AI diminishes the need for a student to develop writing skills suggests that teachers of writing deliver a limited selection of skills that students need to master, such as grammar, punctuation and spelling, all of which are now easily replicated by generative AI. But educators know that teaching is more than just the delivery of a shopping list of skills. Teaching writing is about connection, memory, lived experience, emotion and yes, form and function. AI isn’t going to replace most of those skills.

The AI = skills loss narrative also diminishes students by suggesting that they are lazy or perhaps even antagonistic towards their education, a bunch of chronic cheats who will always try to take the easy way out if we give them the option. Suggesting that students will somehow suddenly forget how to form sentences, just because large language models exist, grossly underestimates the intelligence, creativity and curiosity of almost every young person I’ve ever met.

And now it seems I’m going to contradict myself, because whilst I think the AI equals skills loss argument is problematic, I also think it is entirely correct.

AI Will Lead to Skills Loss

All technological advances lead to skill loss, skill replacement or skill evolution. By now, many people who work with artificial intelligence in education will have heard the comparison drawn with Socrates’ complaints that the invention of writing will lead to the loss of people’s ability to remember information. Writing with a pen and paper, on a stylus, with chalk or with a computer, is a form of technological advancement. It is a shift away from oracy, and it has changed the way we tell stories, interact with one another, and share and record information. Word processors and Microsoft PowerPoint have changed the way that we communicate, something long documented in research extending back to the beginning of computers with graphical user interfaces, and even before then, the introduction of typewriters.

When I took my driver’s license tests in the UK, it was the norm for people to drive manual cars. When I moved to Australia, it was far more commonplace to see people driving automatics, though many people still tested on a manual so that they could potentially drive either. Over the years I taught in Australia, I saw more and more students taking their license with an automatic as manual cars were pushed to the sidelines and predominantly used by people in trade or agricultural settings. I’m sure over the next decade or two, we’ll see an increase in the number of people who don’t bother with driver’s licenses at all, from a combination of improved public transport infrastructures and, in countries where the technology is available, automated vehicles. The skill of driving a manual transmission vehicle will further diminish because of these technological advances.

The more important question is not, will AI replace skills? But, why do we care if AI replaces skills? Honestly, I think a big part of the problem here is nostalgia. We have what seems like a genetic predisposition to think things were better the way they were, but wearing rosy coloured glasses doesn’t make something true.

In the mid 2000s, I remember driving holidays around the UK with my wife using a map that we’d purchased from Morrisons, a chain of supermarkets. It was sparse on detail, missing some roads entirely, but at least had every single Morrisons in the country marked on it, so you could always find a dubious deli sandwich and a petrol station. This was particularly helpful, since I was driving a Vauxhall Corsa which leaked engine oil. And yes, it was a manual transmission.

At some point during those travels, we augmented our navigation technologies with online maps, but this was still pre-smartphones and accessible wireless internet. So this meant loading the entire route on a laptop before setting off, and because we didn’t have a printer, writing the summary of the directions on the back of whatever was handy. The step up from Morrisons map to back-of-the-envelope Google Maps (or MapQuest) directions caused as many problems as it solved, especially when certain key instructions had been missed off entirely, and we ended up driving down some very concerning narrow B roads in the Cotswolds.

Printing directions with mapquest
byu/BaabyBear innostalgia

Eventually, of course, smartphones arrived, and Google Maps moved off the envelope and into an app. Internet became generally more reliable, though still totally inaccessible in the regional parts of Australia where I live now, and navigating became much simpler.

I was taught how to read a map in school. I even remember extracurricular school camps where we did orienteering with a map and a compass. I can still look at a Morrisons map and find the nearest supermarket, but obviously I don’t bother. I use Google Maps. Do I miss navigating by pen and paper? Not at all.

To draw a different generational comparison, I never had to do arithmetic with a log table, because by the time I was going through school, calculators existed. In fact, by the time I hit secondary, we were in the land of scientific calculators and even, gasp, personal computers. I’ve looked at the log tables used by my parents, stored in shoe boxes with aging school reports from the 60s, and honestly, I couldn’t tell you what I’m looking at. Yet I still studied maths all the way through to A levels, and if I absolutely have to, I can still do mental arithmetic.

But it’s not a skill I spent a great deal of my childhood learning, and it’s not one I anticipate needing much in the future. Again, do I lament the fact that I never used a log table because I always had access to a technological solution? Of course not.

We’ve heard all of these arguments before. Spell checkers reduce literacy. Smartphones mean we can’t remember phone numbers any more. YouTube videos on how to DIY fix your car mean that we don’t need mechanics any more… They are, predictably, over simplifications.

Ultimately, we need to take off the rose-coloured glasses and take a long, hard, objective look at which skills we really want to value, and more importantly, why.

Why Do We Care?

The why is the most important part. The why of these skills is going to be the thing that differentiates between an education that students get something out of and one that they’re constantly pushing back against. I think a lot of the fear of AI and skills loss stems from the fact that in the face of these technologies, it’s actually really hard to say why we should continue to teach them. Why should we teach map reading now that GPS and Google Maps exist? Why should we teach mental arithmetic, even though most people have access to incredibly powerful calculators on their phones? Why teach anything at all?

There are many versions of answers to these questions, from the liberal humanist “because it makes you a better person” to the neo-liberal “because it makes you a better citizen, and it’s good for the economy.” There are deep philosophical concerns and arguments that have gone back and forth for thousands of years, and have nothing to do with artificial intelligence and everything to do with what learning is and what education is for.

We’re not going to solve them through heated debates about whether ChatGPT will make everybody illiterate.

But it does raise an important point about broadening our perspectives beyond the educational bubble that we find ourselves in. Although Socrates was concerned about writing 2000 years ago, oracy was still the main vehicle for education in the West until at least the Middle Ages, and for most people far beyond then. The written word and the transfer of knowledge through reading and writing is a relatively new invention in terms of the lineage of the human species and our educational history.

And it was also (and you can argue still is) largely an elitist means of transferring and validating knowledge. The Western idea of knowledge as something which can be taken, pinned down, reproduced and even owned stems from the Eurocentric academic system, which is only a few centuries old, and from religious methods of recording and transmitting knowledge before that.

But other communities around the world have very different approaches to knowledge, different understandings of what it means to know, such as First Nations Australian “ways of knowing, ways of doing, ways of being and ways of valuing“.

That we test the acquisition of knowledge in the West largely through writing and examinations is also a new invention, historically speaking. And if we zoom in even further to standardised approaches to literacy, such as the NAPLAN tests in Australia, curriculum reforms in the US and UK, then we are at the very tip of the history of human experience, limited to decades, not even centuries, let alone millennia of human knowledge sharing.

When I hear somebody say, “AI is a problem because it will lead to students losing the ability to write,” I wonder, what is this hand-wringing over the ability to write? Where does that come from? Why does it exist?

Frankly, as a teacher of writing and as an author, this line of thought comes with a fair amount of existential crisis. When I reflect on my own writing, I never learned to write because of the education system I grew up in. I had some great English teachers, but I was always a reader, and have been writing from a young age outside of the educational context. In recent years, I’ve learned to write in a variety of styles and contexts, from this blog to short fiction, academic writing and the requirements of my PhD. I didn’t learn any of those forms of writing or any of those skills at school other than the basics I was introduced to within the boundaries of the UK national curriculum. I have continued to develop these skills because I want to, and now I incorporate generative artificial intelligence tools into that writing process because I want to.

Maybe this seems like a selfish perspective on skills development, but ultimately, we all do things for our own sake much more readily than because other people tell us to.

Facing the problem head on

The fear that “AI will lead to skills loss” is understandable but misguided. I think a healthier response to this technology would be to acknowledge upfront that it will lead to skills loss, and then ask important questions about why we care:

• Is it because the skills that will be replaced are dear to us personally, and we worry that others won’t get to experience the same knowledge and understanding?
• Is it because we believe those skills are fundamental or vital to the human experience? And if we do believe that, do we need to fact check those thoughts or maybe zoom out beyond the few decades of our experience?

I’m not saying that reading, writing, or any of the other skills that will be displaced and affected by artificial intelligence aren’t important, far from it. But I think we need a more nuanced, less reactionary dialog around what these technologies mean for young people’s development of skills.

If you want to contribute to that dialog, leave a comment here or wherever you found this article.

• LinkedIn
• Bluesky
• Facebook
• ShareCopied to clipboard
• Mastodon
• Threads

Want to learn more about GenAI professional development and advisory services, or just have questions or comments? Get in touch: